Several late stages of the oligodendrocyte (OL) developmental lineage can be identified immunologically in the newborn rat brain. However, OL lineage-specific markers are not available for the detection of the less mature, yet determined, OL precursors. We have developed a retrospective bioassay, combining limiting dilution analysis with a novel culture system, that quantitatively assesses the developmental potential in vivo of phenotypically undefined OL precursors in order to (1) demonstrate their existence, (2) estimate their total number in the premyelinated rat brain, and (3) demonstrate their presence in regions distal to germinal zones at times previously predicted to be devoid of such cells. Between embryonic day (E) 21 and postnatal day (P) 0, cells determined to become oligodendrocytes increase in frequency approximately 5-fold in the whole brain (from one precursor for every 365 cells to 1 in 74), and approximately 2.5-fold in the telencephalon (from 1 in 298 to 1 in 115). From these data it is calculated that a pool of approximately 10(6) phenotypically undefined cells are present in the newborn brain that are able to differentiate into OL in vitro. Further, by applying this assay to tissue samples of subdomains of the developing cerebellum, we have demonstrated that such cells are present in large numbers as early as E20 in regions sparsely populated with cells expressing the blastic neural cell marker ganglioside GD3, suggesting that they migrated to this position as a pre-GD3-expressing cell. These results significantly change the predicted ontogeny of the oligodendrocyte lineage and should fuel the ongoing search for these early OL precursors.